The Impacts of Climate Change on Human Health - Amazon Simple ...

THE IMPACTS OF CLIMATE CHANGE ON HUMAN HEALTH IN THE UNITED STATES

A Scientific Assessment

2

TEMPERATURE-RELATED DEATH AND ILLNESS Lead Authors Marcus C. Sarofim* U.S. Environmental Protection Agency Shubhayu Saha Centers for Disease Control and Prevention Michelle D. Hawkins National Oceanic and Atmospheric Administration David M. Mills Abt Associates Contributing Authors Jeremy Hess University of Washington Radley Horton Columbia University Patrick Kinney Columbia University Joel Schwartz Harvard University Alexis St. Juliana Abt Associates

Recommended Citation: Sarofim, M.C., S. Saha, M.D. Hawkins, D.M. Mills, J. Hess, R. Horton, P. Kinney, J. Schwartz, and A. St. Juliana, 2016: Ch. 2: Temperature-Related Death and Illness. The Impacts of Climate Change on Human Health in the United States: A Scientific Assessment. U.S. Global Change Research Program, Washington, DC, 43–68. http://dx.doi.org/10.7930/J0MG7MDX

On the web: health2016.globalchange.gov

*Chapter Coordinator

U.S. Global Change Research Program

2

TEMPERATURE-RELATED DEATH AND ILLNESS

FPO

Key Findings Future Increases in Temperature-Related Deaths Key Finding 1: Based on present-day sensitivity to heat, an increase of thousands to tens of thousands of premature heat-related deaths in the summer [Very Likely, High Confidence] and a decrease of premature coldrelated deaths in the winter [Very Likely, Medium Confidence] are projected each year as a result of climate change by the end of the century. Future adaptation will very likely reduce these impacts (see Changing Tolerance to Extreme Heat Finding). The reduction in cold-related deaths is projected to be smaller than the increase in heat-related deaths in most regions [Likely, Medium Confidence].

Even Small Differences from Seasonal Average Temperatures Result in Illness and Death Key Finding 2: Days that are hotter than usual in the summer or colder than usual in the winter are both associated with increased illness and death [Very High Confidence]. Mortality effects are observed even for small differences from seasonal average temperatures [High Confidence]. Because small temperature differences occur much more frequently than large temperature differences, not accounting for the effect of these small differences would lead to underestimating the future impact of climate change [Likely, High Confidence].

Changing Tolerance to Extreme Heat Key Finding 3: An increase in population tolerance to extreme heat has been observed over time [Very High Confidence]. Changes in this tolerance have been associated with increased use of air conditioning, improved social responses, and/or physiological acclimatization, among other factors [Medium Confidence]. Expected future increases in this tolerance will reduce the projected increase in deaths from heat [Very Likely, Very High Confidence].

Some Populations at Greater Risk Key Finding 4: Older adults and children have a higher risk of dying or becoming ill due to extreme heat [Very High Confidence]. People working outdoors, the socially isolated and economically disadvantaged, those with chronic illnesses, as well as some communities of color, are also especially vulnerable to death or illness [Very High Confidence].

44

2–TEMPERATURE-RELATED DEATH AND ILLNESS

2.1 Introduction The Earth is warming due to elevated concentrations of greenhouse gases, and will continue to warm in the future. U.S. average temperatures have increased by 1.3°F to 1.9°F since record keeping began in 1895, heat waves have become more frequent and intense, and cold waves have become less frequent across the nation (see Ch. 1: Introduction). Annual average U.S. temperatures are projected to increase by 3°F to 10°F by the end of this century, depending on future emissions of greenhouse gases and other factors.1 These temperature changes will have direct effects on human health. Days that are hotter than the average seasonal temperature in the summer or colder than the average seasonal temperature in the winter cause increased levels of illness and death by compromising the body’s ability to regulate its temperature or by inducing direct or indirect health complications. Figure 1 provides a conceptual model of the various climate drivers, social factors, and environmental and institutional factors that

can interact to result in changes in illness and deaths as a result of extreme heat. Increasing concentrations of greenhouse gases lead to an increase of both average and extreme temperatures, leading to an increase in deaths and illness from heat and a potential decrease in deaths from cold. Challenges involved in determining the temperature–death relationship include a lack of consistent diagnoses on death certificates and the fact that the health implications of extreme temperatures are not absolute, differing from location to location and changing over time. Both of these issues can be partially addressed through the use of statistical methods. Climate model projections of future temperatures can be combined with the estimated relationships between temperatures and health in order to assess how deaths and illnesses resulting from temperature could change in the future. The impact of a warming climate on deaths and illnesses will not be realized equally as a number of populations, such as children, the elderly, and economically disadvantaged groups, are especially vulnerable to temperature.

Climate Change and Health—Extreme Heat

Figure 1: This conceptual diagram illustrates the key pathways by which climate change influences human health during an extreme heat event, and potential resulting health outcomes (center boxes). These exposure pathways exist within the context of other factors that positively or negatively influence health outcomes (gray side boxes). Key factors that influence vulnerability for individuals are shown in the right box, and include social determinants of health and behavioral choices. Key factors that influence vulnerability at larger scales, such as natural and built environments, governance and management, and institutions, are shown in the left box. All of these influencing factors can affect an individual’s or a community’s vulnerability through changes in exposure, sensitivity, and adaptive capacity and may also be affected by climate change. See Chapter 1: Introduction for more information. U.S. Global Change Research Program

45

Impacts of Climate Change on Human Health in the United States


2.2 Contribution of Extreme Temperatures to Death and Illness Temperature extremes most directly affect health by compromising the body’s ability to regulate its internal temperature. Loss of internal temperature control can result in a cascade of illnesses, including heat cramps, heat exhaustion, heatstroke, and hyperthermia in the presence of extreme heat, and hypothermia and frostbite in the presence of extreme cold. Temperature extremes can also worsen chronic conditions such as cardiovascular disease, respiratory disease, cerebrovascular disease, and diabetes-related conditions. Prolonged exposure to high temperatures is associated with increased hospital admissions for cardiovascular, kidney, and respiratory disorders. Exposures to high minimum temperatures may also reduce the ability of the human body to recover from high daily maximum temperatures.

2.3 Defining Temperature Exposures

thresholds that are absolute. For example, one combined measure of humidity and temperature is known as the wet bulb temperature. As the wet bulb temperature reaches or exceeds the threshold of 35°C (95°F), the human body can no longer cool through perspiration, and recent evidence suggests that there is a physical heat tolerance limit in humans to sustained temperatures above 35°C that is similar across diverse climates.13 The combined effects of temperature and humidity have been incorporated in tools such as heat index tables, which reflect how combinations of heat and relative humidity “feel.” The heat index in these tools is often presented with notes about the potential nature and type of health risks different combinations of temperature and humidity may pose, along with confounding conditions such as exposure to direct sunlight or strong winds. Variations in heat wave definitions make it challenging to compare results across studies or determine the most appropriate public health warning systems.8, 14 This is important as the associations between deaths and illnesses and extreme heat conditions vary depending on the methods used for defining the extreme conditions.2, 15, 16

Extreme temperatures are typically defined by some measure, for example, an ambient temperature, heat index (a combination of temperature and humidity), or wind chill (a combination of temperature and wind speed), exceeding predefined thresholds over a number of days.2, 3, 4, 5, 6, 7, 8 Extremes can be 2.4 Measuring the Health Impact of Temperature defined by average, minimum, or maximum daily temperatures, by nighttime temperatures, or by daytime temperaTwo broad approaches are used to study the relationship betures. However, there is no standard method for defining a tween temperatures and illness and death: direct attribution heat wave or cold wave. There are dramatic differences in and statistical methods.17, 18 the observed relationships between temperature, death, and Direct Attribution Studies illness across different regions and seasons; these relationships vary based on average temperatures in those locations With direct attribution, researchers link health outcomes to and the timing of the heat or cold event. For example, a 95°F temperatures based on assigned diagnosis codes in medical day in Vermont will have different implications for health records such as hospital admissions and death certificates. than a 95°F day in Texas, and For example, the International similarly, a 95°F day in May will Classification of Diseases (ICDTemperature extremes most directly affect have different implications than 10) contains specific codes for health by compromising the body's ability to 9, 10, 11, 12 one in August (this is attributing deaths to exposure regulate its internal temperature. further discussed in"Evidence to excessive natural heat (X30) of Adaptation to Temperature and excessive natural cold Extremes" on page 49). There(X31).19 However, medical fore, in some cases, temperature extremes are defined by records will not include information on the weather conditions comparison to some local average (for example, the top 1% of at the time of the event or preceding the event. It is generally warmest days recorded in a particular location) rather than to accepted that direct attribution underestimates the number of some absolute temperature (such as 95°F). While temperature people who die from temperature extremes. Reasons for this extremes are generally determined based on weather stainclude difficulties in diagnosing heat-related and cold-related tion records, the exposure of individuals will depend on their deaths, lack of consistent diagnostic criteria, and difficulty in location: urban heat islands, microclimates, and differences identifying, or lack of reporting, heat or cold as a factor that between indoor and outdoor temperatures can all lead to worsened a preexisting medical condition.9, 17 Heat-related differences between weather station data and actual expodeaths are often not reported as such if another cause of sure. The indoor environment is particularly important as most death exists and there is no well-publicized heat wave. An people spend the majority of their time inside. additional challenging factor in deaths classified as X31 (cold) deaths is that a number of these deaths result from situations One exception to using relative measures of temperature is involving substance use/abuse and/or contact with water, that there are some critical physical and weather condition both of which can contribute to hypothermia.20, 21


46



Statistical Studies

exacerbating the cause of death, only recording the ultimate cause, such as a stroke or a heart attack (see, for example, Figure 2, where the excess deaths during the 1995 Chicago heat wave clearly exceeded the number of deaths recorded as heat-related on death certificates). Statistical methods focus on determining how temperature contributes to premature deaths and illness and therefore are not susceptible to this kind of undercount, though they face potential biases due to time-varying factors like seasonality. Both methods depend on temperatures measured at weather stations, though the actual temperature exposure of individuals may differ. In short, while the focus on temperature is consistent in both methods, the methods potentially evaluate very different combinations of deaths and weather conditions.

Statistical studies measure the impact of temperature on death and illness using methods that relate the number of cases (for example, total daily deaths in a city) to observed weather conditions and other socio-demographic factors. These statistical methods determine whether the temperature conditions were associated with increased deaths or illness above longer-term average levels. These associations establish the relationship between temperature and premature deaths and illness. In some cases, particularly with extreme temperature conditions, the increase in premature deaths and illness can be quite dramatic and the health impact may be referred to in terms of excess deaths or illnesses. Methods for evaluating the impact of temperature in these models vary. Many studies include all the days in the study period, which makes it possible to capture changes in deaths resulting from small variations of temperatures from their seasonal averages. Other methods restrict the analysis to days that exceed some threshold for extreme heat or cold conditions.22 Some studies incorporate methods that determine different health relationships for wind, air pressure, and cloud cover as well as the more common temperature and humidity measures.15 Another approach is to identify a heat event and compare observed illness and deaths during the event with a carefully chosen comparison period.23, 24, 25 Many of these methods also incorporate socio-demographic factors (for example, age, race, and poverty) that may affect the temperature–death relationship. Comparing Results of Direct Attribution and Statistical Studies Comparing death estimates across studies is therefore complicated by the use of different criteria for temperature extremes, different analytical methods, varying time periods, and different affected populations. Further, it is widely accepted that characteristics of extreme temperature events such as duration, intensity, and timing in season directly affect actual death totals.2, 12 Estimates of the average number of deaths attributable to heat and cold considering all temperatures, rather than just those associated with extreme events, provide an alternative for considering the mortality impact of climate change.26, 27 Statistical studies can also offer insights into what aspects of a temperature extreme are most important. For example, there are indications that the relationship between high nighttime temperatures and mortality is more pronounced than the relationship for daytime temperatures.12, 16 These two methods (direct attribution and statistical approaches) yield very different results for several reasons. First, statistical approaches generally suggest that the actual number of deaths associated with temperature is far greater than those recorded as temperature-related in medical records. Medical records often do not capture the role of heat in U.S. Global Change Research Program

2.5 Observed Impact of Temperature on Deaths A number of extreme temperature events in the United States have led to dramatic increases in deaths, including events in Kansas City and St. Louis in 1980, Philadelphia in 1993, Chicago in 1995, and California in 2006. (See Figure 2 for more on the July 1995 heat wave in Chicago).28, 29, 30, 31, 32 Recent U.S. studies in specific communities and for specific extreme temperature events continue to conclude that extreme temperatures, particularly extreme heat, result in premature deaths.7, 30, 36, 37 This finding is further reinforced by a growing suite of regional- and national-scale studies documenting an increase in deaths following extreme temperature conditions, using both direct attribution17 and statistical approaches.9, 10, 12, 15, 38 The connection between heat events and deaths is also evident internationally. The European heat wave of 2003 is an especially notable example, as it is estimated to have been responsible for between 30,000 and 70,000 premature deaths.39 However, statistical approaches find that elevated death rates are seen even for less extreme temperatures. These approaches find an optimal temperature, and show that there are more deaths at any temperatures that are higher or lower than that optimal temperature.11, 40 Even though the increase in deaths per degree are smaller near the optimum than at more extreme temperatures, because the percentage of days that do not qualify as extreme are large,41 it can be important to address the changes in deaths that occur for these smaller temperature differences. A recent analysis of U.S. deaths from temperature extremes based on death records found an average of approximately 1,300 deaths per year from 2006 to 2010 coded as resulting from extreme cold exposures, and 670 deaths per year coded as resulting from exposure to extreme heat.17 These results, and those from all similar studies that rely solely on coding within medical records to determine cause of deaths, will underestimate the actual number of deaths due to extreme temperatures.17, 42 For example, some statistical approaches estimate that more than 1,300 deaths per year in the United

47



Heat-Related Deaths in Chicago in the Summer of 1995 Figure 2 illustrates an example of excess deaths following an extreme heat event. In this case, excess deaths are determined by calculating the difference between daily observed deaths in Chicago during the worst of the heat wave (starting on July 11) and longerterm daily averages for this time of year. The period of extreme heat extended from June 21 through August 10, 1995. Research into the event suggests it was the combination of high humidity, high daily maximum temperatures, and high daily minimum temperatures that made this event truly exceptional.33 This event

is estimated to have resulted in nearly 700 excess deaths in Chicago, based on a statistical approach.35 By comparison, a direct attribution approach based on death certificates found only 465 deaths were attributed to extreme heat during this time period.29 This kind of underestimate resulting from relying on death certificates is common. It is reasonable to expect that deaths may be even less likely to be attributed to extreme heat during a heat wave that, unlike the Chicago event, does not receive a great deal of public attention.

Heat-Related Deaths During the 1995 Chicago Heat Wave

Figure 2: This figure shows the relationship between high temperatures and deaths observed during the 1995 Chicago heat wave. The large spike in deaths in mid-July of 1995 (red line) is much higher than the average number of deaths during that time of year (orange line), as well as the death rate before and after the heat wave. This increase in the rate of deaths occurred during and after the heat wave, as shown here by temperatures exceeding 100°F during the day (green line). Humidity and high nighttime temperatures were also key contributing factors to this increase in deaths.33 The number of excess deaths has been estimated to be about 700 based on statistical methods, but only 465 deaths in Cook County were classified as “heat-related” on death certificates during this same period,29 demonstrating the tendency of direct attribution to undercount total heat-related deaths. (Figure source: EPA 2014)34


48



States are due to extreme heat.15, 43 Different approaches to attributing cause of death lead to differences in the relative number of deaths attributed to heat and cold.44 Studies based on statistical approaches have found that, despite a larger number of deaths being coded as related to extreme cold rather than extreme heat, and a larger mortality rate in winter overall, the relationship between mortality and an additional day of extreme heat is generally much larger than the relationship between mortality and an additional day of extreme cold.12 Confounding Factors and Effect Modifiers While the direct attribution approach underestimates the number of deaths resulting from extreme temperature events, there are a few ways in which the statistical approach may lead to an overestimation. However, any overestimation due to these potential confounding factors and effect modifiers is thought to be much smaller than the direct attribution underestimation.12

The relationship between mortality and an additional day of extreme heat is generally much larger than the relationship between mortality and an additional day of extreme cold.

Evidence of Adaptation to Temperature Extremes

The impact on human health of a given temperature event (for example, a 95°F day) can depend on where and when it occurs. The evidence also shows larger changes in deaths and The first potential overestimation results from the connection hospitalizations in response to elevated temperatures in cities between elevated temperatures and other variables that corwhere temperatures are typically cooler as compared with relate with temperature, such as poor air quality. This connecwarmer cities.9, 11, 40, 51, 52 This suggests that people can adapt, tion involves a combination of factors, including stagnant air at least partially, to the average temperature that they are masses and changes in the atmospheric chemistry that affect used to experiencing. Some of this effect can be explained by the concentrations of air pollutants such as ozone or particudifferences in infrastructure. For example, locations with highlate matter (see Ch. 3: Air Qualer average temperature, such ity Impacts). If some portion of as the Southeast, will generally the deaths during extreme heat have greater prevalence and use The impact on human health of a given events are actually a result of temperature event (for example, a 95°day) can of air conditioning. However, the higher levels of atmospheric there is also evidence that there depend on where and when it occurs. pollution that are correlated is a physiological acclimatization with these events, then includ(the ability to gradually adapt ing those deaths in a statistical to heat), with changes in sweat analysis to determine the relationship of increased heat on volume and timing, blood flow and heat transfer to the skin, human health would result in double counting deaths.10, 45, 46, 47 and kidney function and water conservation occurring over However, this issue is often addressed by including air polluthe course of weeks to months of exposure to a hot climate.53 26 tion and other correlated variables in statistical modeling. For example, as a result of this type of adaptation, heat events later in the summer have less of an impact on deaths than A second consideration when using statistical approaches to those earlier in the summer, all else being equal,15 although determine the relationship between temperature and deaths some of this effect is also due to the deaths of some of the is whether some of the individuals who died during the temmost vulnerable earlier in the season. However, children and perature event were already near death, and therefore the older adults remain vulnerable given their reduced ability to temperature event could be considered to have “displaced” regulate their internal temperature and limited acclimatization the death by a matter of days rather than having killed a percapacities.53 son not otherwise expected to die. This effect is referred to as mortality displacement. There is still no consensus regarding An increased tolerance to extreme temperatures has also the influence of mortality displacement on premature death been observed over multiyear and multidecadal periods.9, 10, 54, 55, 56 estimates, but this effect generally accounts for a smaller porThis improvement is likely due to some combination of 7, 12, tion of premature deaths as events become more extreme. physiological acclimatization, increased prevalence and use of 48, 49, 50 air conditioning,10 and general improvements in public health over time,9, 54 but the relative importance of each is not yet clear.56


49



Recent changes in urban planning and development programs reflect an adaptive trend implemented partially in response to the anticipated temperature health risks of climate change. For example, because urban areas tend to be warmer than surrounding rural areas (the “urban heat island” effect), there is an increased emphasis on incorporating green space and other technologies, such as cool roofs, in new development or redevelopment projects.57 Similarly, programs that provide advice and services in preparation for or response to extreme temperatures continue to increase in number and expand the scope of their activity (see for example guidance documents on responses to extreme temperature developed by the Centers for Disease Control and Prevention and the Environmental Protection Agency).58, 59 Continued changes in personal behavior as a result of these efforts, for example, seeking access to air-conditioned areas during extreme heat events or limiting outside activity, may continue to change future exposure to extreme temperatures and other climate-sensitive health stressors, such as outdoor air pollutants and vectors for disease such as ticks or mosquitoes. Observed Trends in Heat Deaths As discussed in Chapter 1, U.S. average temperature has increased by 1.3°F to 1.9°F since 1895, with much of that increase occurring since 1970, though this temperature increase has not been uniform geographically and some regions, such as the Southeast, have seen little increase in temperature and extreme heat over time.1, 15 This warming is attributable to elevated concentrations of greenhouse gases and it has been estimated that three-quarters of moderately hot extremes are already a result of this historical warming.60 As discussed in the previous section, there have also been changes in the tolerance of populations within the United States to extreme temperatures. Changes in mortality due to high temperatures are therefore a result of the combination of higher temperatures and higher heat tolerance. Use of the direct attribution approach, based on diagnosis codes in medical records, to examine national trends in heat mortality over time is challenging because of changes in classification methods over time.34

Certain occupational groups such as agricultural workers, construction workers, and electricity and pipeline utility workers are at increased risk for heat- and cold-related illness, especially where jobs involve heavy exertion. U.S. Global Change Research Program

The few studies using statistical methods that have presented total mortality estimates over time suggest that, over the last several decades, reductions in mortality due to increases in tolerance have outweighed increases in mortality due to increased temperatures.15, 61

2.6 Observed Impact of Temperature on Illness Temperature extremes are linked to a range of illnesses reported at emergency rooms and hospitals. However, estimates for the national burden of illness associated with extreme temperatures are limited. Using a direct attribution approach, an analysis of a nationally representative database from the Healthcare Utilization Project (HCUP) produced an annual average estimate of 65,299 emergency visits for acute heat illness during the summer months (May through September)—an average rate of 21.5 visits for every 100,000 people each year.62 This result was based only on recorded diagnosis codes for hyperthermia and probably underestimates the true number of heat-related healthcare visits, as a wider range of health outcomes is potentially affected by extreme heat. For example, hyperthermia is not the only complication from extreme heat, and not every individual that suffers from a heat illness visits an emergency department. In a national study of Medicare patients from 2004 to 2005, an annual average of 5,004 hyperthermia cases and 4,381 hypothermia cases were reported for inpatient and outpatient visits.63 None of these studies link health episodes to observed temperature data, thus limiting the opportunity to attribute these adverse outcomes to specific heat events or conditions. High ambient heat has been associated with adverse impacts for a wide range of illnesses.25 Examples of illnesses associated with extreme heat include cardiovascular, respiratory, and renal illnesses; diabetes; hyperthermia; mental health issues; and preterm births. Children spend more time outdoors and have insufficient ability for physiologic adaptation, and thus may be particularly vulnerable during heat waves.64 Respiratory illness among the elderly population was most commonly reported during extreme heat.65 Statistical studies examine the association between extreme heat and illness using data from various healthcare access points (such as hospital admissions, emergency department visits, and ambulance dispatches). The majority of these studies examine the association of extreme heat with cardiovascular and respiratory illnesses. For these particular health outcomes, the evidence is mixed, as many studies observed elevated risks of illness during periods of extreme heat but others found no evidence of elevated levels of illness.24, 51, 66, 67, 68, 69, 70 The evidence on some of the other health outcomes is more robust. Across emergency department visits and hospital admissions, high temperature have been associated with renal diseases, electrolyte imbalance, and hyperthermia.24, 67,

50



These health risks vary not only across types of illness but also for the same illness across different healthcare settings. In general, evidence for associations with morbidity outcomes, other than cardiovascular impacts, is strong. 71, 72

While there is still uncertainty about how levels of heat-related illnesses are expected to change with projected increases in summer temperature from climate change,41 advances have been made in surveillance of heat-related illness. For example, monitoring of emergency ambulance calls during heat waves can be used to establish real-time surveillance systems to identify extreme heat events.73 The increase in emergency visits for a wide range of illnesses during the 2006 heat wave in California points to the potential for using this type of information in real-time health surveillance systems.24

2.7 Projected Deaths and Illness from Temperature Exposure Climate change will increase the frequency and severity of future extreme heat events while also resulting in generally warmer summers and milder winters,1 with implications for human health. Absent further adaptation, these changes are expected to lead to an increase in illness and death from increases in heat, and reductions in illness and death resulting from decreases in cold, due to changes in outcomes such as heat stroke, cardiovascular disease, respiratory disease, cerebrovascular disease, and kidney disorders.41, 74 A warmer future is projected to lead to increases in future mortality on the order of thousands to tens of thousands of additional premature deaths per year across the United States by the end of this century.22, 38, 54, 75, 76, 77, 78, 79 Studies differ in which regions of the United States are examined and in how they account for factors such as adaptation, mortality displacement, demographic changes, definitions of heat waves and extreme cold, and air quality factors, and some studies examine only extreme events while others take into account the health effects of smaller deviations from average seasonal temperatures. Despite these differences there is reasonable agreement on the magnitude of the projected changes. Additionally, studies have projected an increase in premature deaths due to increases in temperature for Chicago, IL,39, 80 Dallas, TX,18 the Northeast corridor cities of Boston, MA, New York, NY, and Philadelphia, PA,18, 26, 81, 82 Washington State,83, 84 California,85 or a group of cities including Portland, OR; Minneapolis and St. Paul, MN; Chicago, IL; Detroit, MI; Toledo, Cleveland, Columbus, and Cincinnati, OH; Pittsburgh and Philadelphia, PA; and Washington, DC.86 However, these regional projections use a variety of modeling strategies and therefore show more variability in mortality estimates than studies that are national in scope. Less is known about how non-fatal illnesses will change in response to projected increases in heat. However, hospital admissions related to respiratory, hormonal, urinary, genital, and U.S. Global Change Research Program

renal problems are generally projected to increase.72, 87 Kidney stone prevalence has been linked to high temperatures, possibly due to dehydration leading to concentration of the salts that form kidney stones. In the United States, an increased rate of kidney stones is observed in southern regions of the country, especially the Southeast. An expansion of the regions where the risk of kidney stones is higher is therefore plausible in a warmer future.88, 89, 90 The decrease in deaths and illness due to reductions in winter cold have not been as well studied as the health impacts of increased heat, but the reduction in premature deaths from cold are expected to be smaller than the increase in deaths from heat in the United States.22, 26, 38, 41, 75, 77 While this is true nationally (with the exception of Barreca 2012),75 it may not hold for all regions within the country.27 Similarly, international studies have generally projected a net increase in deaths from a warming climate, though in some regions, decreases in cold mortality may outweigh increases in heat mortality.91 The projected net increase in deaths is based in part on historical studies that show that an additional extreme hot day leads to more deaths than an additional extreme cold day, and in part on the fact that the decrease in extreme cold deaths is limited as the total number of cold deaths approaches zero in a given location. It is important to distinguish between generally higher wintertime mortality rates that are not strongly associated with daily temperatures—such as respiratory infections and some cardiovascular disease 12, 92—from mortality that is more directly related to the magnitude of the cold temperatures. Some recent studies have suggested that factors leading to higher wintertime mortality rates may not be sensitive to climate warming, and that deaths due to these factors are expected to occur with or without climate change. Considering this, some estimates of wintertime mortality may overstate the benefit of climate change in reducing wintertime deaths.49, 93, 94 The U.S. population has become less sensitive to heat over time. Factors that have contributed to this change include infrastructure improvements, including increased access and use of air conditioning in homes and businesses, and improved societal responses, including increased access to public health programs and healthcare.15, 54, 61, 95, 96, 97 Projecting these trends into the future is challenging, but this trend of increasing tolerance is projected to continue, with future changes in adaptive capacity expected to reduce the future increase in mortality.56 However, there are limits to adaptation, whether physiological53 or sociotechnical (for example, air conditioning, awareness programs, or cooling centers). While historically adaptation has outpaced warming, most studies project a future increase in mortality even when including assumptions regarding adaptation.18, 22, 81, 85, 91 Additionally, the occurrence of events such as power outages simultaneous with a heat wave may reduce some of these adaptive benefits. Such simultaneous events can

51



Research Highlight: Modeling the Effect of Warming on U.S. Deaths Importance: A warming climate is expected to result in more days that are warmer than today’s usual temperature in the summer, leading to an increase in heat-related deaths. A warming climate is also expected to result in fewer days that are colder than today’s usual temperatures in the winter, leading to a decrease in cold-related deaths. Understanding these changes is an important factor in understanding the human health response to climate change. Objective: A quantitative projection of future deaths from heat and cold for 209 U.S. cities with a total population of over 160 million inhabitants. Method: A relationship between average daily temperature and deaths by city and month was developed using historical data on deaths and temperatures from 1996–2006, generating results for both same-day temperature and the average of the previous five-day temperatures to account for delayed responses to temperature. Cities, which are defined using county borders, were allocated to nine different clusters based on similarity of climates. Temperature–death relationships were refined for cities within a given cluster based on the other cities in that cluster. Projections of temperature in future time periods were based on the RCP6.0 scenario from two climate models: the Geophysical Fluid Dynamic Laboratory–Coupled Physical Model 3 (GFDL–CM3) and the Model for Interdisciplinary Research on Climate (MIROC5). These projections were adjusted to match the historical data from the same weather stations that were used in the statistical analysis. Further details can be found in Schwartz et al. 2015.27

Projected Changes in Temperature-Related Death Rates

Figure 3: This figure shows the projected decrease in death rates due to warming in colder months (October–March, top left), the projected increase in death rates due to warming in the warmer months (April–September, top right), and the projected net change in death rates (combined map, bottom), comparing results for 2100 to those for a 1990 baseline period in 209 U.S. cities. These results are from one of the two climate models (GFDL–CM3 scenario RCP6.0) studied in Schwartz et al. (2015). In the study, mortality data for a city is based on county-level records, so the borders presented reflect counties corresponding to the study cities. Geographic variation in the death rates are due to a combination of differences in the amount of projected warming and variation in the relationship between deaths and temperatures derived from the historical health and temperature data. These results are based on holding the 2010 population constant in the analyses, with no explicit assumptions or adjustment for potential future adaptation. Therefore, these results reflect only the effect of the anticipated change in climate over time. (Figure source: Schwartzet al. 2015)27


52



Research Highlight: Modeling the Effect of Warming on U.S. Deaths, continued Results: The modeling done for this study projects that future warming, without any adjustments for future adaptation, will lead to an increase in deaths during hotter months, defined as April–September, and a decrease in deaths during colder months, defined as October–March. Overall, this leads to a total net increase of about 2,000 to 10,000 deaths per year in the 209 cities by the end of the century compared to a 1990 baseline (Figure 4). Net effects vary from city to city, and a small number of cities are projected to experience a decrease in deaths (Figures 3 and 4). Conclusions: This study is an improvement on previous studies because it examines a greater proportion of the U.S. population, uses more recent data on deaths, takes advantage of similar relationships between deaths and temperature between nearby cities to generate more statistically robust results, and addresses the difference in these relationships by month of the year. The results are consistent with most of the previous studies in projecting that climate change will lead to an increase in heat deaths on the order of thousands to tens of thousands of annual deaths by the end of the century compared to the 1990 baseline, and that the increase in deaths from heat will be larger than the reduction in deaths from cold. In contrast to some previous similar studies,22 some individual cities show a net reduction in future deaths due to future warming, mainly in locations where the population is already well-adapted to heat but poorly prepared for cold (like Florida). Barreca 201275 also shows net mortality benefits in some counties, though with a different spatial pattern due to humidity effects. Some other studies also have different spatial patterns, projecting high excess mortality in Southern states despite a lower risk per degree change, due to larger increases in frequency and duration of heat waves in that region.79 Like most previous studies, this analysis does not account for the effects of further adaptation on future mortality. Results are based on the temperature–death relationships observed for the period from 1996 to 2006, which reflect historical adaptation to extreme temperatures. However, future adaptation would, all else equal, mean that these results may overestimate the potential impact of climate change on changes in both heat- and cold-related deaths. This study increases the confidence in the key finding that the number of heat deaths will increase in the future compared to a future with no climate change, and that the increase in heat deaths will be larger than the reduction in cold deaths.

Projected Changes in Deaths in U.S. Cities by Season Figure 4: This figure shows the projected increase in deaths due to warming in the summer months (hot season, April–September), the projected decrease in deaths due to warming in the winter months (cold season, October–March), and the projected net change in deaths for the 209 U.S. cities examined. These results compare projected deaths for future reporting years to results for the year 1990 while holding the population constant at 2010 levels and without any quantitative adjustment for potential future adaptation, so that temperature–death relationships observed in the last decade of the available data (1997–2006) are assumed to remain unchanged in projections over the 21st century. With these assumptions, the figure shows an increasing health benefit in terms of reduced deaths during the cold season (October–March) over the 21st century from warming temperatures, while deaths during the hot season (April–September) increase. Overall, the additional deaths from the warming in the hot season exceed the reduction in deaths during the cold season, resulting in a net increase in deaths attributable to temperature over time as a result of climate change. The baseline and future reporting years are based on 30-year periods where possible, with the exception of 2100: 1990 (1976–2005), 2030 (2016–2045), 2050 (2036–2065), and 2100 (2086–2100). (Figure source: adapted from Schwartz et al. 2015)27


53



be more common because of the additional demand on the electricity grid due to high air-conditioning usage.98 Another potential effect is that if current trends of population growth and migration into large urban areas continue, there may be an increasing urban heat island effect which will magnify the rate of warming locally, possibly leading to more heat-related deaths and fewer cold-related deaths. Projected changes in future health outcomes associated with extreme temperatures can be difficult to quantify. Projections can depend on 1) the characterization of population sensitivity to temperature event characteristics such as magnitude, duration, and humidity; 2) differences in population sensitivity depending on the timing and location of an extreme event; 3) future changes in baseline rates of death and illness as well as human tolerance and adaptive capacity; 4) the changing proportions of vulnerable populations, including the elderly, in the future; and 5) uncertainty in climate projections.

2.8 Populations of Concern for Death and Illness from Extreme Temperatures Impacts of temperature extremes are geographically varied and disproportionally affect certain populations of concern (see also Ch. 9: Populations of Concern).41 Certain populations are more at risk for experiencing detrimental consequences of exposure to extreme temperatures due to their sensitivity to hot and cold temperatures and limitations to their capacity for adapting to new climate conditions. Older adults are a rapidly growing population in the United States, and heat impacts are projected to occur in places where older adults are heavily concentrated and therefore most exposed.99 Older adults are at higher risk for temperature-related mortality and morbidity, particularly those who have preexisting diseases, those who take certain medications that affect thermoregulation or block nerve impulses (for example, beta-blockers, major tranquilizers, and diuretics), those who are living alone, or those with limited mobility (see also Ch. 9: Populations of Concern).17, 24, 42, 45, 100 The relationship between increased temperatures and death in older adults is well-understood with strong evidence of heat-related vulnerability for adults over 65 and 75 years old.101 An increased risk for respiratory and cardiovascular death is observed in older adults during temperature extremes due to reduced thermoregulation.17, 42, 45, 65 Morbidity studies have also identified links between increased temperatures and respiratory and cardiovascular hospitalizations in older adults.65 Children are particularly vulnerable because they must rely on others to help keep them safe. This is especially true in environments that may lack air conditioning, including homes, schools, or cars (see also Ch. 9: Populations of Concern).102 The primary health complications observed in children exposed to extreme heat include dehydration, electrolyte imbalance, fever, renal disease, heat stress, and hyperthermia.64 InfecU.S. Global Change Research Program

FPO

Physiological factors and participation in vigorous outdoor activities make children particularly vulnerable to extreme heat.

tious and respiratory diseases in children are affected by both hot and cold temperatures.64 Inefficient thermoregulation, reduced cardiovascular output, and heightened metabolic rate are physiological factors driving vulnerability in children to extreme heat. Children also spend a considerable amount of time outdoors and participating in vigorous physical activities.17, 42, 64, 103 High-school football players are especially vulnerable to heat illness (see also Ch. 9: Populations of Concern).104 A limited number of studies show evidence of cold-related mortality in children. However, no study has examined the relationship between cold temperature and cause-specific mortality.64 Pregnant women are also vulnerable to temperature extremes as preterm birth has been associated with extreme heat.42, 105, 106 Elevated heat exposure can increase dehydration, leading to the release of labor-inducing hormones.107 Extreme heat events are also associated with adverse birth outcomes, such as low birth weight and infant mortality (see Ch. 9: Populations of Concern). Where a person lives, works, or goes to school can also make them more vulnerable to health impacts from extreme temperatures. Of particular concern for densely populated cities is the urban heat island effect, where manmade surfaces absorb sunlight during the day and then radiate the stored energy at night as heat. This process will exacerbate any warming from climate change and limit the potential relief of cooler nighttime temperatures in urban areas.81 In addition to the urban heat island effect, land cover characteristics and poor air quality combine to increase the impacts of high ambient temperatures for city dwellers and further increase the burden on populations of concern within the urban area.12, 17, 45, 108 The homeless are often more exposed to heat and cold extremes, while also sharing many risk factors with other populations of concern such as social isolation, psychiatric illness, and other health issues.109 Race, ethnicity, and socioeconomic status can affect vulnerability to temperature extremes. Non-Hispanic Black persons

54



have been identified as being more vulnerable than other racial and ethnic groups to detrimental consequences of exposure to temperature extremes.17, 42, 45, 103, 110, 111 One study found that non-Hispanic Blacks were 2.5 times more likely to experience heat-related mortality compared to non-Hispanic Whites, and non-Hispanic Blacks had a two-fold risk of dying from a heat-related event compared to Hispanics.17 Evidence of racial differences in heat tolerance due to genetic differences is inconclusive.110 However, other factors may contribute to increased vulnerability of Black populations, including comorbidities (co-existing chronic conditions) that increase susceptibility to higher temperatures, disparities in the availability and use of air conditioning and in heat risk-related land cover characteristics (for example, living in urban areas prone to heat-island effects), and environmental justice issues.17, 42, 108, 110, 112 Overall, the link between temperature extremes, race, ethnicity, and socioeconomic status is multidimensional and dependent on the outcome being studied. Education level, income, safe housing, occupational risks, access to health care, and baseline health and nutrition status can further distort the association between temperature extremes, race, and ethnicity.45, 110 Outdoor workers spend a great deal of time exposed to temperature extremes, often while performing vigorous activities. Certain occupational groups such as agricultural workers, construction workers, and electricity and pipeline utility workers are at increased risk for heat- and cold-related illness, especially where jobs involve heavy exertion.100, 113, 114 One study found failure of employers to provide for acclimatization to be the factor most clearly associated with heat-related death in workers.113 Mental, behavioral, and cognitive disorders can be triggered or exacerbated by heat waves. Specific illnesses impacted by heat include dementia, mood disorders, neurosis and stress, and substance abuse.100, 115, 116, 117 Some medications interfere with thermoregulation, thereby increasing vulnerability to heat.116 One study in Australia found that hospital admissions for mental and behavioral disorders increased by 7.3% during heat waves above 80°F.115 Studies have also linked extreme heat and increased aggressive behavior. (See also Ch. 8: Mental Health).

2.9 Emerging and Cross-Cutting Issues Emerging and cross-cutting issues include 1) disparate ways that extreme temperature and health are related, 2) urban and rural differences, 3) interactions between impacts and future changes in adaptation, and 4) projections of extreme temperature events. The health effects addressed in this chapter are not the only ways in which heat and health are related. For example, research indicates that hotter temperatures may lead to an increase in violent crime118 and could negatively affect the labor U.S. Global Change Research Program

force, especially occupational health for outdoor sectors.119, 120 Extreme temperatures also interact with air quality, which can complicate estimating how extreme temperature events impact human health in the absence of air quality changes (see "Confounding Factors and Effect Modifiers" on page 49). In addition, increased heat may also increase vulnerability to poor air quality and allergens, leading to potential non-linear health outcome responses. Extreme temperature events, as well as other impacts from climate change, can also be associated with changes in electricity supply and distribution that can have important implications for the availability of heating and air conditioning, which are key adaptive measures. Though the estimates of the health impact from extreme heat discussed in the “Research Highlight” were produced only for urban areas (which provided a large sample size for statistical validity), there is also emerging evidence regarding high rates of heat-related illness in rural areas.6, 62 Occupational exposure and a lack of access to air conditioning are some of the factors that may make rural populations particularly susceptible to extreme heat. There are quantitative challenges to using statistical methods to estimate mortality impacts of temperatures in rural areas due to lower population density and more dispersed weather stations, but rural residents have also demonstrated vulnerability to heat events.121 Other changes in human behavior will also have implications for the linkage between climate and heat-related illness. Changes in building infrastructure as a response to changes in temperature can have impacts on indoor air quality. Similarly, changes in behavior as a result of temperature changes, for example, seeking access to air conditioning, can change exposure to indoor and outdoor pollution and vectorborne diseases (see Ch. 3: Air Quality Impacts; Ch. 5: Vectorborne Diseases). Finally, projecting climate variability and the most extreme temperature events can be more challenging than projecting average warming. Extreme temperatures may rise faster than average temperatures,122 with the coldest days warming faster than average for much of the twentieth century, and the warmest days warming faster than average temperatures in the past 30 years.123 Extremely high temperatures in the future may also reach levels outside of past experience, in which case statistically based relationships may no longer hold for those events. There have been suggestive links between rapid recent Arctic sea ice loss124 and an increased frequency of cold125 and warm extremes,126 but this is an active area of research with conflicting results.127, 128 In regions where temperature variability increases, mortality will be expected to increase; mortality is expected to decrease in regions where variability decreases.129

55



2.10 Research Needs In addition to the emerging issues identified above, the authors highlight the following potential areas for additional scientific and research activity on temperature-related illness and death based on their review of the literature. Improved modeling and more robust projections of climate variability and extreme temperatures will enhance the modeling of health impacts associated with extremes of heat and cold. While the surveillance for temperature-related deaths is relatively robust, understanding the impacts of future changes in heat waves and extreme temperatures can be improved with better surveillance and documentation of non-fatal illnesses, including hospitalizations and emergency room visits, for temperature-associated reasons. With growing implementation of heat early warning systems around the country, there is also a need for the development of evaluation methods and associated collection of data to be able to assess effectiveness of such systems and other means of health adaptation. Future assessments can benefit from research activities that: • further explore the associations between exposure to a range of high and low temperatures and exacerbation of illnesses across locations and healthcare settings; • improve understanding of how genetic factors and social determinants contribute to vulnerability to illness and death from extreme temperature exposures; • analyze the combined health effects of temperature and other discrete climate-sensitive stressors, such as changing air quality, smoke from wildfires, or impacts of extreme weather events; • attribute changes in observed mortality to a changing climate; • develop effective adaptive responses to reduce the potential adverse health outcomes attributable to changing temperatures; and • explore how future adaptive measures and behaviors can be included in quantitative models of health impacts associated with extreme temperatures


56


Supporting Evidence The chapter was developed through technical discussions of relevant evidence and expert deliberation by the report authors at several workshops, teleconferences, and email exchanges. The authors considered inputs and comments submitted by the public, the National Academies of Sciences, and Federal agencies. For additional information on the overall report process, see Appendices 2 and 3.

supported this finding. Of those studies that examine both heat and cold at the national scale, only Barreca found that the reductions in cold deaths would more than compensate for the increase in heat deaths.22, 27, 38, 75, 77 Barreca’s study was novel in terms of its treatment of humidity, finding that weather that was both cold and dry, or both hot and humid, was associated with higher mortality. However, this treatment of humidity was not the cause of the difference with other studies, as leaving out humidity actually showed a greater benefit from future climate change. Instead, the author stated that the reduction in net deaths was a result of relying on counties with over 100,000 inhabitants, and that using a state-level model covering all U.S. deaths would lead to a prediction of an increase of 1.7% in mortality rates rather than a decrease of 0.1%. The finding by the majority of studies at a national scale that heat deaths will increase more than cold deaths will decrease is consistent with studies at smaller spatial scales.26 Moreover, several studies provide rationales for why heat mortality is expected to outpace cold mortality,12, 22, 27 and some studies suggest that cold mortality may not be responsive to warming.49, 93, 94 Barnett et al. (2012) showed that cold waves were not generally associated with an increase in deaths beyond the mortality already associated with cold weather, in contrast to heat waves.2

The content of this chapter was determined after reviewing the collected literature. The authors determined that there was substantial literature available to characterize both observed and projected mortality from elevated temperatures, with sufficient literature available to also characterize mortality from cold as well as cold-related hospitalizations and illness. Populations of concern were also considered to be a high priority for this chapter. As discussed in the chapter, there were limitations in terms of the state of the literature on understanding how future adaptation will influence climaterelated changes in temperature-related mortality, addressing the impact of temperature on rural populations, and examining health-related endpoints other than mortality and morbidity. KEY FINDING TRACEABLE ACCOUNTS

Future Increases in Temperature-Related Deaths Key Finding 1: Based on present-day sensitivity to heat, an increase of thousands to tens of thousands of premature heat-related deaths in the summer [Very Likely, High Confidence] and a decrease of premature cold-related deaths in the winter [Very Likely, Medium Confidence] are projected each year as a result of climate change by the end of the century. Future adaptation will very likely reduce these impacts (see Changing Tolerance to Extreme Heat Finding). The reduction in cold-related deaths is projected to be smaller than the increase in heat-related deaths in most regions [Likely, Medium Confidence].

Major uncertainties The largest remaining uncertainties concern questions of future adaptation, which are discussed in Key Finding #3. A related uncertainty involves the link between the temperatures measured at weather stations and the temperatures experienced by individuals. As long as the relationship between the weather station and the microclimate or indoor/outdoor difference remains constant, this should not impair projections. However, as microclimates, building construction, or behavior change, the relationship between recorded weather station temperature and actual temperature exposure will change. This is related to, but broader than, the question of adaptation. Additionally, there are uncertainties regarding the non-linearities of heat response with increasing temperatures.

Description of evidence base An extensive literature examines projections of mortality due to increasing temperatures. In particular, nine studies were identified that provide heat mortality projections in the United States for at least 10% of the U.S. population.22, 27, 38, 54, 75, 76, 77, 78, 79 Each of these studies projected an increase in heatrelated mortality due to projections of future warming, though several noted the potential modification effect of adaptation (discussed in Key Finding #3). In general, the magnitude of projected increases in annual premature deaths in these studies was in the hundreds to thousands by mid-century, and thousands to tens of thousands by the end of the century, when scaled to the total U.S. population. These conclusions are further supported by studies at the city, county, and state level.18, 26, 39, 80, 81, 82, 83, 84, 85

Assessment of confidence and likelihood based on evidence There is high confidence that heat deaths will very likely increase in the future compared to a future without climate change, based on high agreement and a large number of studies as well as consistency across scenarios and regions. Because there are fewer studies on winter mortality, and because studies exist that suggest that winter mortality is not strongly linked to temperatures, there is medium confidence that deaths due to extreme cold will very likely decrease. The majority of the studies that examine both heat and cold deaths find that the increase in heat deaths due to climate change will likely be larger than the decrease in cold deaths in most regions, but there are a limited number of such studies, leading to an assessment of medium confidence.

The Third National Climate Assessment (2014 NCA) found that “While deaths and injuries related to cold events are projected to decline due to climate change, these reductions are not expected to compensate for the increase in heat-related deaths,”41 and studies published since that time have further U.S. Global Change Research Program

57



Even Small Differences from Seasonal Average Temperatures Result in Illness and Death

Major uncertainties This finding reflects consideration of a number of recent studies17, 54 not referenced in the recent 2014 NCA.41 There is a consensus of studies linking extreme temperatures and mortality, and a growing body of literature demonstrating that smaller differences in temperature are also linked with mortality. However, the mortality displacement effect, and the fact that deaths that do not occur during an identified heat wave are less likely to be directly attributed to extreme heat, contribute to continuing uncertainty about the magnitude of the effect of temperature on mortality.

Key Finding 2: Days that are hotter than usual in the summer or colder than usual in the winter are both associated with increased illness and death [Very High Confidence]. Mortality effects are observed even for small differences from seasonal average temperatures [High Confidence]. Because small temperature differences occur much more frequently than large temperature differences, not accounting for the effect of these small differences would lead to underestimating the future impact of climate change [Likely, High Confidence].

Assessment of confidence and likelihood based on evidence There is very high confidence in the relationship between extreme temperatures and premature deaths due to the consistency and strength of the literature, particularly given the different study designs that produce this result. There is high confidence that small temperature deviations from normal temperatures contribute to premature mortality due to high agreement among those studies that have examined the issue. Though some studies indicate that for these small temperature differences, mortality displacement may play a larger role than for more extreme temperatures. Fewer studies have examined the role of these smaller temperature differences in projections, but the directionality of the effect is clear, so the determination of the authors was that not including this effect would likely lead to an underestimate of future mortality, with high confidence.

Description of evidence base Two well-recognized conclusions from the literature are that extreme temperatures lead to illness and premature death and that these extreme temperatures are best described in relation to local average seasonal temperatures rather than absolute temperature values. Epidemiological studies find an increase in mortality at temperatures that are high related to the local average.9, 10, 12, 15, 17, 38 Based on absolute temperatures, Anderson and Bell 2011 found that cities in the South and Southeast were the least sensitive to heat, demonstrating acclimatization.9 Illness has been linked with hot daily average temperature4, 6, 51, 69, 71 and apparent temperature, among other metrics.3, 66, 68, 87 Across studies, adverse health episodes were most strongly associated with exposures to high temperatures that occurred on the same day or the previous day.3, 51 However, a cumulative effect of heat was also observed at periods of up to one week after exposure, tapering off beyond seven days.69, 105 Cardiovascular and respiratory illness has been most commonly examined in relation to extreme heat, but the association is more varied for illness than for mortality due to effects across age groups69, 70 and differences in morbidity risk associated with emergency room records versus hospital admissions.4, 6, 24, 51, 66, 67, 68, 69, 70

Changing Tolerance to Extreme Heat Key Finding 3: An increase in population tolerance to extreme heat has been observed over time [Very High Confidence]. Changes in this tolerance have been associated with increased use of air conditioning, improved social responses, and/or physiological acclimatization, among other factors [Medium Confidence]. Expected future increases in this tolerance will reduce the projected increase in deaths from heat [Very Likely, Very High Confidence].

The evidence for mortality is clearest for extreme temperatures, as addressed in threshold-based studies,12 but studies that account for smaller changes in temperature found mortality changes even for small deviations of temperature.11, 27 This is consistent with studies showing a U-shaped relationship of temperature and mortality—while there may be some plateau near the optimal temperature, the plateau is often small, and not always coincident with the seasonal average temperature.11, 40 However, some of the individuals who die in response to elevated temperatures were already near death, and so the temperature event is sometimes considered to have “displaced” the death by a matter of days rather than created an additional death. Studies have found that this effect is generally below 50% of the total deaths, and is much smaller than that (10% or less) for the most extreme events, such as the 2003 European heat wave.12, 48, 49, 50 In contrast, one recent study found that in seven U.S. cities mortality displacement was greater than 80% for small temperature deviations and around 50% even for the 3% of warmest events in the study sample.7


Description of evidence base The increasing tolerance of the U.S. population to extreme heat has been shown by a number of studies.9, 10, 54 However, there is less confidence in attributing this increase in tolerance: increased prevalence and use of air conditioning, physiological adaptation, available green space, and improved social responses have all been proposed as explanatory factors. There have been some indications (Sheridan et al. 2009)97 that tolerance improvements in the United States might have plateaued, but Bobb et al. 2014 found continuing improvements through 2005.54 Several approaches to including adaptation have been used in temperature mortality projection studies. For example, two studies used an “analog city” approach, where the response of the population to future temperatures in a given city is assumed to be equal to that of a city with a hotter present-day climate.22, 81 Another approach is to assume that critical temperature thresholds change by

58



some quantity over time.18, 91 A third approach is to calculate sensitivity to air conditioning prevalence in the present, and make assumptions about air conditioning in the future.85 In general, inclusion of adaptation limits the projected increase in deaths, sometimes modestly, other times dramatically. However, approaches used to account for adaptation may be optimistic. Historically, adaptive measures have occurred as a response to extreme events, and therefore could be expected to lag warming.39, 96 While the increase in mortality projected in these studies is reduced, the studies generally found that mortality increases compared to present day even under optimistic adaptation assumptions.18, 22, 81, 85 A limit to adaptation may be seen in that even in cities with nearly 100% air conditioning penetration, heat deaths are observed today.

rate. Children also spend a considerable amount of time outdoors and participating in vigorous activities.17, 42, 64, 103 A limited number of studies found evidence of cold-related mortality in children; however, no study has examined the relationship between cold temperature and cause-specific mortality.64

Major uncertainties While studies have been published in recent years that include adaptation in sensitivity analyses,22 this remains a challenging area of research. Difficulties in attributing observed increases in tolerance make it challenging to project future changes in tolerance, whether due to autonomous adaptation by individuals or planned adjustments by governments. Extrapolation of acclimatization is limited as there must be an increase in temperature beyond which acclimatization will not be possible.

Race, ethnicity, and socioeconomic status have been shown to impact vulnerability to temperature extremes. Several studies have identified non-Hispanic Black populations to be more vulnerable than other racial and ethnic groups for experiencing detrimental consequences of exposure to temperature extremes.17, 42, 45, 103, 110 Studies suggest comorbidities that enhance susceptibility to higher temperatures, availability and use of air conditioning, disparities in heat risk-related land cover characteristics, and other environmental justice issues contribute to increased vulnerability of non-Hispanic Blacks.17, 42, 108, 110, 112

Assessment of confidence and likelihood based on evidence There is very high confidence that a decrease in sensitivity to heat events has occurred based on high agreement between studies, but only medium confidence that this decrease is due to some specific combination of air conditioning prevalence, physiological adaptation, presence of green space, and improved social responses because of the challenges involved in attribution. There is very high confidence that mortality due to heat will very likely be reduced compared to a noadaptation scenario when adaptation is included, because all studies examined were in agreement with this conclusion, though the magnitude of this reduction is poorly constrained.

Dementia, mood disorders, neurosis and stress-related illnesses, and substance abuse are shown to be impacted by extreme heat.100, 115, 116, 117 Some medications interfere with thermoregulation, increasing vulnerability to heat.116

Certain occupational groups that spend a great deal of time exposed to extreme temperatures, such as agricultural workers, construction workers, and electricity and pipeline utility workers, are at increased risk for heat- and cold-related illness, especially where jobs involve heavy exertion.100, 113, 114 Lack of heat-illness-prevention programs in the workplace that include provisions for acclimatization was found to be a factor strongly associated with extreme temperature-related death.113

Major uncertainties The literature available at the time of the development of the 2014 NCA had identified a number of vulnerable populations that were disproportionately at risk during heat waves, and literature since that time has only strengthened the understanding of the elevated risks for these populations. There continues to be a need for better understanding of the relative importance of genetics and environmental justice issues with regards to the observed higher risk for nonHispanic Blacks, more work on understanding the risks to pregnant women from extreme temperature events, and a better understanding of the relationship between extreme cold vulnerabilities in populations of concern.

Some Populations at Greater Risk Key Finding 4: Older adults and children have a higher risk of dying or becoming ill due to extreme heat [Very High Confidence]. People working outdoors, the socially isolated and economically disadvantaged, those with chronic illnesses, as well as some communities of color, are also especially vulnerable to death or illness [Very High Confidence].

Assessment of confidence and likelihood based on evidence Although some details regarding causation and identifying the most vulnerable subpopulations still require research, there is a large body of literature that demonstrates the increased vulnerability to extreme heat of a number of groups, and therefore there is very high confidence that the listed populations of concern are at greater risk of temperaturerelated death and illness.

Description of evidence base The relationship between increased temperatures and deaths in elderly populations is well-understood. An increased risk of respiratory and cardiovascular death is observed in elderly populations during temperature extremes due to reduced thermoregulation.17, 42, 45, 65 Studies cite dehydration, electrolyte imbalance, fever, heat stress, hyperthermia, and renal disease as the primary health conditions in children exposed to heat waves. Causes of heatrelated illness in children include inefficient thermoregulation, reduced cardiovascular output, and heightened metabolic U.S. Global Change Research Program

59



DOCUMENTING UNCERTAINTY This assessment relies on two metrics to communicate the degree of certainty in Key Findings. See Appendix 4: Documenting Uncertainty for more on assessments of likelihood and confidence.

Confidence Level

Likelihood

Very High

Very Likely

Strong evidence (established theory, multiple sources, consistent results, well documented and accepted methods, etc.), high consensus High Moderate evidence (several sources, some consistency, methods vary and/or documentation limited, etc.), medium consensus

≥ 9 in 10 Likely ≥ 2 in 3 As Likely As Not

Medium Suggestive evidence (a few sources, limited consistency, models incomplete, methods emerging, etc.), competing schools of thought

≈ 1 in 2 Unlikely ≤ 1 in 3 Very Unlikely ≤ 1 in 10

Low Inconclusive evidence (limited sources, extrapolations, inconsistent findings, poor documentation and/or methods not tested, etc.), disagreement or lack of opinions among experts

PHOTO CREDITS Pg. 43–Construction worker: © Fotosearch Pg. 44–Large Crowd: © iStockImages.com/Ints Vikmanis Pg. 49–Snowstorm: © iStockImages.com/Dreef Pg. 50–Construction worker: © Fotosearch Pg. 54–Young baseball catcher: © iStockImages.com/jpbcpa


60


References 1.

2.

3.

4.

Walsh, J., D. Wuebbles, K. Hayhoe, J. Kossin, K. Kunkel, G. Stephens, P. Thorne, R. Vose, M. Wehner, J. Willis, D. Anderson, S. Doney, R. Feely, P. Hennon, V. Kharin, T. Knutson, F. Landerer, T. Lenton, J. Kennedy, and R. Somerville, 2014: Ch. 2: Our changing climate. Climate Change Impacts in the United States: The Third National Climate Assessment. Melillo, J.M., T.C. Richmond, and G.W. Yohe, Eds. U.S. Global Change Research Program, Washington, D.C., 19-67. http://dx.doi.org/10.7930/J0KW5CXT Barnett, A.G., S. Hajat, A. Gasparrini, and J. Rocklöv, 2012: Cold and heat waves in the United States. Environmental Research, 112, 218-224. http://dx.doi.org/10.1016/j. envres.2011.12.010 Gronlund, C.J., A. Zanobetti, J.D. Schwartz, G.A. Wellenius, and M.S. O’Neill, 2014: Heat, heat waves, and hospital admissions among the elderly in the United States, 1992– 2006. Environmental Health Perspectives, 122, 1187-1192. http://dx.doi.org/10.1289/ehp.1206132 Lavigne, E., A. Gasparrini, X. Wang, H. Chen, A. Yagouti, M.D. Fleury, and S. Cakmak, 2014: Extreme ambient temperatures and cardiorespiratory emergency room visits: Assessing risk by comorbid health conditions in a time series study. Environmental Health, 13, 5. http://dx.doi. org/10.1186/1476-069x-13-5

9.

Anderson, G.B. and M.L. Bell, 2011: Heat waves in the United States: Mortality risk during heat waves and effect modification by heat wave characteristics in 43 U.S. communities. Environmental Health Perspectives, 119, 210-218. http://dx.doi.org/10.1289/ehp.1002313

10. Guo, Y., A.G. Barnett, and S. Tong, 2012: High temperatures-related elderly mortality varied greatly from year to year: Important information for heat-warning systems. Scientific Reports, 2. http://dx.doi.org/10.1038/srep00830 11. Lee, M., F. Nordio, A. Zanobetti, P. Kinney, R. Vautard, and J. Schwartz, 2014: Acclimatization across space and time in the effects of temperature on mortality: A time-series analysis. Environmental Health, 13, 89. http://dx.doi. org/10.1186/1476-069X-13-89 12. Medina-Ramón, M. and J. Schwartz, 2007: Temperature, temperature extremes, and mortality: A study of acclimatisation and effect modification in 50 US cities. Occupational and Environmental Medicine, 64, 827-833. http://dx.doi. org/10.1136/oem.2007.033175 13. Sherwood, S.C. and M. Huber, 2010: An adaptability limit to climate change due to heat stress. Proceedings of the National Academy of Sciences, 107, 9552-9555. http://dx. doi.org/10.1073/pnas.0913352107

5.

Lin, S., M. Luo, R.J. Walker, X. Liu, S.-A. Hwang, and R. Chinery, 2009: Extreme high temperatures and hospital admissions for respiratory and cardiovascular diseases. Epidemiology, 20, 738-746. http://dx.doi.org/10.1097/ EDE.0b013e3181ad5522

14. Kent, S.T., L.A. McClure, B.F. Zaitchik, T.T. Smith, and J.M. Gohlke, 2014: Heat waves and health outcomes in Alabama (USA): The importance of heat wave definition. Environmental Health Perspectives, 122, 151–158. http://dx.doi. org/10.1289/ehp.1307262

6.

Lippmann, S.J., C.M. Fuhrmann, A.E. Waller, and D.B. Richardson, 2013: Ambient temperature and emergency department visits for heat-related illness in North Carolina, 2007-2008. Environmental Research, 124, 35-42. http://dx. doi.org/10.1016/j.envres.2013.03.009

15. Kalkstein, L.S., S. Greene, D.M. Mills, and J. Samenow, 2011: An evaluation of the progress in reducing heat-related human mortality in major US cities. Natural Hazards, 56, 113-129. http://dx.doi.org/10.1007/s11069-010-9552-3

7.

Saha, M.V., R.E. Davis, and D.M. Hondula, 2014: Mortality displacement as a function of heat event strength in 7 US cities. American Journal of Epidemiology, 179, 467-474. http://dx.doi.org/10.1093/aje/kwt264

8.

Smith, T.T., B.F. Zaitchik, and J.M. Gohlke, 2013: Heat waves in the United States: Definitions, patterns and trends. Climatic Change, 118, 811-825. http://dx.doi.org/10.1007/ s10584-012-0659-2


16. Zhang, K., R.B. Rood, G. Michailidis, E.M. Oswald, J.D. Schwartz, A. Zanobetti, K.L. Ebi, and M.S. O’Neill, 2012: Comparing exposure metrics for classifying ‘dangerous heat’ in heat wave and health warning systems. Environment International, 46, 23-29. http://dx.doi.org/10.1016/j. envint.2012.05.001 17. Berko, J., D.D. Ingram, S. Saha, and J.D. Parker, 2014: Deaths Attributed to Heat, Cold, and Other Weather Events in the United States, 2006–2010. National Health Statistics Reports No. 76, July 30, 2014, 15 pp. National Center for Health Statistics, Hyattsville, MD. http://www.cdc.gov/ nchs/data/nhsr/nhsr076.pdf

61



18. Gosling, S.N., J.A. Lowe, G.R. McGregor, M. Pelling, and B.D. Malamud, 2009: Associations between elevated atmospheric temperature and human mortality: A critical review of the literature. Climatic Change, 92, 299-341. http://dx. doi.org/10.1007/s10584-008-9441-x 19. WHO, 2004: International Statistical Classification of Diseases and Related Health Problems, 10th Revision (ICD–10), 2nd ed. World Health Organization, Geneva, Switzerland. http://www.who.int/classifications/icd/ICD-10_2nd_ed_ volume2.pdf 20. CDC, 2005: Extreme Cold: A Prevention Guide to Promote Your Personal Health and Safety. 13 pp. U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, Atlanta, GA. http://www.bt.cdc.gov/disasters/winter/pdf/extreme-cold-guide.pdf 21. CDC, 2006: Hypothermia-related deaths – United States, 1999-2002 and 2005. MMWR. Morbidity and Mortality Weekly Report, 55, 282-284. http://www.cdc.gov/mmwr/ preview/mmwrhtml/mm5510a5.htm 22. Mills, D., J. Schwartz, M. Lee, M. Sarofim, R. Jones, M. Lawson, M. Duckworth, and L. Deck, 2015: Climate change impacts on extreme temperature mortality in select metropolitan areas in the United States. Climatic Change, 131, 83-95. http://dx.doi.org/10.1007/s10584-014-1154-8 23. Bustinza, R., G. Lebel, P. Gosselin, D. Bélanger, and F. Chebana, 2013: Health impacts of the July 2010 heat wave in Quebec, Canada. BMC Public Health, 13, 56. http://dx.doi. org/10.1186/1471-2458-13-56

27. Schwartz, J.D., M. Lee, P.L. Kinney, S. Yang, D. Mills, M. Sarofim, R. Jones, R. Streeter, A. St. Juliana, J. Peers, and R.M. Horton, 2015: Projections of temperature-attributable premature deaths in 209 U.S. cities using a cluster-based Poisson approach. Environmental Health, 14. http://dx.doi. org/10.1186/s12940-015-0071-2 28. CDC, 1994: Heat-related deaths--Philadelphia and United States, 1993-1994. MMWR. Morbidity and Mortality Weekly Report, 43, 453-455. http://www.cdc.gov/mmwr/preview/ mmwrhtml/00031773.htm 29. CDC, 1995: Heat-related mortality – Chicago, July 1995. MMWR. Morbidity and Mortality Weekly Report, 44, 577-579. http://www.cdc.gov/mmwr/preview/mmwrhtml/00038443.htm 30. Hoshiko, S., P. English, D. Smith, and R. Trent, 2010: A simple method for estimating excess mortality due to heat waves, as applied to the 2006 California heat wave. International Journal of Public Health, 55, 133-137. http://dx.doi. org/10.1007/s00038-009-0060-8 31. Jones, T.S., A.P. Liang, E.M. Kilbourne, M.R. Griffin, P.A. Patriarca, S.G. Fite Wassilak, R.J. Mullan, R.F. Herrick, D. Donnell, Jr., K. Choi, and S.B. Thacker, 1982: Morbidity and mortality associated with the July 1980 heat wave in St Louis and Kansas City, Mo. JAMA: The Journal of the American Medical Association, 247, 3327–3331. http://dx.doi. org/10.1001/jama.1982.03320490025030 32. Jones, S., M. Griffin, A. Liang, and P. Patriarca, 1980: The Kansas City Heat Wave, July 1980: Effects of Health, Preliminary Report. Centers for Disease Control, Atlanta, GA.

24. Knowlton, K., M. Rotkin-Ellman, G. King, H.G. Margolis, D. Smith, G. Solomon, R. Trent, and P. English, 2009: The 2006 California heat wave: Impacts on hospitalizations and emergency department visits. Environmental Health Perspectives, 117, 61-67. http://dx.doi.org/10.1289/ehp.11594

33. Karl, T.R. and R.W. Knight, 1997: The 1995 Chicago heat wave: How likely is a recurrence? Bulletin of the American Meteorological Society, 78, 1107-1119. http://dx.doi.org/10 .1175/1520-0477(1997)078%3C1107:tchwhl%3E2.0.co;2

25. Ye, X., R. Wolff, W. Yu, P. Vaneckova, X. Pan, and S. Tong, 2012: Ambient temperature and morbidity: A review of epidemiological evidence. Environmental Health Perspectives, 120, 19-28. http://dx.doi.org/10.1289/ehp.1003198

34. EPA, 2014: Climate Change Indicators in the United States, 2014. 3rd edition. EPA 430-R-14-04, 107 pp. U.S. Environmental Protection Agency, Washington, D.C. http://www. epa.gov/climatechange/pdfs/climateindicators-full-2014.pdf

26. Li, T., R.M. Horton, and P.L. Kinney, 2013: Projections of seasonal patterns in temperature-related deaths for Manhattan, New York. Nature Climate Change, 3, 717-721. http:// dx.doi.org/10.1038/nclimate1902

35. Kaiser, R., A. Le Tertre, J. Schwartz, C.A. Gotway, W.R. Daley, and C.H. Rubin, 2007: The effect of the 1995 heat wave in Chicago on all-cause and cause-specific mortality. American Journal of Public Health, 97, S158-S162. http:// dx.doi.org/10.2105/ajph.2006.100081


62



36. Harlan, S.L., J.H. Declet-Barreto, W.L. Stefanov, and D.B. Petitti, 2013: Neighborhood effects on heat deaths: Social and environmental predictors of vulnerability in Maricopa County, Arizona. Environmental Health Perspectives, 121, 197-204. http://dx.doi.org/10.1289/ehp.1104625

44. Dixon, P.G., D.M. Brommer, B.C. Hedquist, A.J. Kalkstein, G.B. Goodrich, J.C. Walter, C.C. Dickerson, S.J. Penny, and R.S. Cerveny, 2005: Heat mortality versus cold mortality: A study of conflicting databases in the United States. Bulletin of the American Meteorological Society, 86, 937-943. http:// dx.doi.org/10.1175/bams-86-7-937

37. Madrigano, J., M.A. Mittleman, A. Baccarelli, R. Goldberg, S. Melly, S. von Klot, and J. Schwartz, 2013: Temperature, myocardial infarction, and mortality: Effect modification by individual- and area-level characteristics. Epidemiology, 24, 439-446. http://dx.doi.org/10.1097/ EDE.0b013e3182878397

45. Anderson, B.G. and M.L. Bell, 2009: Weather-related mortality: How heat, cold, and heat waves affect mortality in the United States. Epidemiology, 20, 205-213. http://dx.doi. org/10.1097/EDE.0b013e318190ee08

38. Deschênes, O. and M. Greenstone, 2011: Climate change, mortality, and adaptation: Evidence from annual fluctuations in weather in the US. American Economic Journal: Applied Economics, 3, 152-185. http://dx.doi.org/10.1257/ app.3.4.152

46. Analitis, A., P. Michelozzi, D. D’Ippoliti, F. de’Donato, B. Menne, F. Matthies, R.W. Atkinson, C. Iñiguez, X. Basagaña, A. Schneider, A. Lefranc, A. Paldy, L. Bisanti, and K. Katsouyanni, 2014: Effects of heat waves on mortality: Effect modification and confounding by air pollutants. Epidemiology, 25, 15-22. http://dx.doi.org/10.1097/ EDE.0b013e31828ac01b

39. Hayhoe, K., S. Sheridan, L. Kalkstein, and S. Greene, 2010: Climate change, heat waves, and mortality projections for Chicago. Journal of Great Lakes Research, 36, 65-73. http:// dx.doi.org/10.1016/j.jglr.2009.12.009 40. Gasparrini, A., Y. Guo, M. Hashizume, E. Lavigne, A. Zanobetti, J. Schwartz, A. Tobias, S. Tong, J. Rocklöv, B. Forsberg, M. Leone, M. De Sario, M.L. Bell, Y.-L.L. Guo, C.-f. Wu, H. Kan, S.-M. Yi, M. de Sousa Zanotti Stagliorio Coelho, P.H.N. Saldiva, Y. Honda, H. Kim, and B. Armstrong, 2015: Mortality risk attributable to high and low ambient temperature: A multicountry observational study. The Lancet, 386, 369-375. http://dx.doi.org/10.1016/ S0140-6736(14)62114-0

47. Madrigano, J., D. Jack, G.B. Anderson, M.L. Bell, and P.L. Kinney, 2015: Temperature, ozone, and mortality in urban and non-urban counties in the Northeastern United States. Environmental Health, 14, 3. http://dx.doi. org/10.1186/1476-069X-14-3 48. Kalkstein, L.S., 1998: Climate and human mortality: Relationships and mitigating measures. Advances in Bioclimatology, 5, 161-177. http://dx.doi.org/10.1007/978-3-64280419-9_7 49. Kinney, P.L., M. Pascal, R. Vautard, and K. Laaidi, 2012: Winter mortality in a changing climate: Will it go down? Bulletin Epidemiologique Hebdomadaire, 12-13, 5-7.

41. Luber, G., K. Knowlton, J. Balbus, H. Frumkin, M. Hayden, J. Hess, M. McGeehin, N. Sheats, L. Backer, C.B. Beard, K.L. Ebi, E. Maibach, R.S. Ostfeld, C. Wiedinmyer, E. Zielinski-Gutiérrez, and L. Ziska, 2014: Ch. 9: Human health. Climate Change Impacts in the United States: The Third National Climate Assessment. Melillo, J.M., T.C. Richmond, and G.W. Yohe, Eds. U.S. Global Change Research Program, Washington, D.C., 220-256. http://dx.doi.org/10.7930/J0PN93H5

50. Le Tertre, A., A. Lefranc, D. Eilstein, C. Declercq, S. Medina, M. Blanchard, B. Chardon, P. Fabre, L. Filleul, J.-F. Jusot, L. Pascal, H. Prouvost, S. Cassadou, and M. Ledrans, 2006: Impact of the 2003 heatwave on all-cause mortality in 9 French cities. Epidemiology, 17, 75-79. http://dx.doi. org/10.1097/01.ede.0000187650.36636.1f

42. Basu, R. and J.M. Samet, 2002: Relation between elevated ambient temperature and mortality: A review of the epidemiologic evidence. Epidemiologic Reviews, 24, 190-202. http://dx.doi.org/10.1093/epirev/mxf007

51. Anderson, G.B., F. Dominici, Y. Wang, M.C. McCormack, M.L. Bell, and R.D. Peng, 2013: Heat-related emergency hospitalizations for respiratory diseases in the Medicare population. American Journal of Respiratory and Critical Care Medicine, 187, 1098-103. http://dx.doi.org/10.1164/ rccm.201211-1969OC

43. Vanos, J.K., L.S. Kalkstein, and T.J. Sanford, 2015: Detecting synoptic warming trends across the US midwest and implications to human health and heat-related mortality. International Journal of Climatology, 35, 85-96. http://dx. doi.org/10.1002/joc.3964


63



52. Zanobetti, A., M.S. O’Neill, C.J. Gronlund, and J.D. Schwartz, 2012: Summer temperature variability and longterm survival among elderly people with chronic disease. Proceedings of the National Academy of Sciences, 109, 66086613. http://dx.doi.org/10.1073/pnas.1113070109

62. Hess, J.J., S. Saha, and G. Luber, 2014: Summertime acute heat illness in U.S. emergency departments from 2006 through 2010: Analysis of a nationally representative sample. Environmental Health Perspectives, 122, 1209-1215. http:// dx.doi.org/10.1289/ehp.1306796

53. Hanna, E.G. and P.W. Tait, 2015: Limitations to thermoregulation and acclimatization challenge human adaptation to global warming. International Journal of Environmental Research and Public Health, 12, 8034-8074. http://dx.doi. org/10.3390/ijerph120708034

63. Noe, R.S., J.O. Jin, and A.F. Wolkin, 2012: Exposure to natural cold and heat: Hypothermia and hyperthermia medicare claims, United States, 2004–2005. American Journal of Public Health, 102, e11-e18. http://dx.doi.org/10.2105/ ajph.2011.300557

54. Bobb, J.F., R.D. Peng, M.L. Bell, and F. Dominici, 2014: Heat-related mortality and adaptation to heat in the United States. Environmental Health Perspectives, 122, 811-816. http://dx.doi.org/10.1289/ehp.1307392

64. Xu, Z., R.A. Etzel, H. Su, C. Huang, Y. Guo, and S. Tong, 2012: Impact of ambient temperature on children’s health: A systematic review. Environmental Research, 117, 120-131. http://dx.doi.org/10.1016/j.envres.2012.07.002

55. Petkova, E.P., A. Gasparrini, and P.L. Kinney, 2014: Heat and mortality in New York City since the beginning of the 20th century. Epidemiology, 25, 554-560. http://dx.doi. org/10.1097/ede.0000000000000123

65. Åström, D.O., F. Bertil, and R. Joacim, 2011: Heat wave impact on morbidity and mortality in the elderly population: A review of recent studies. Maturitas, 69, 99-105. http://dx.doi.org/10.1016/j.maturitas.2011.03.008

56. Hondula, D.M., R.C. Balling, Jr., J.K. Vanos, and M. Georgescu, 2015: Rising temperatures, human health, and the role of adaptation. Current Climate Change Reports, 1, 144154. http://dx.doi.org/10.1007/s40641-015-0016-4

66. Basu, R., D. Pearson, B. Malig, R. Broadwin, and R. Green, 2012: The effect of high ambient temperature on emergency room visits. Epidemiology, 23, 813-820. http://dx.doi. org/10.1097/EDE.0b013e31826b7f97

57. Stone, B.J., J. Vargo, P. Liu, D. Habeeb, A. DeLucia, M. Trail, Y. Hu, and A. Russell, 2014: Avoided heat-related mortality through climate adaptation strategies in three US cities. PLoS ONE, 9, e100852. http://dx.doi.org/10.1371/ journal.pone.0100852

67. Green, R.S., R. Basu, B. Malig, R. Broadwin, J.J. Kim, and B. Ostro, 2010: The effect of temperature on hospital admissions in nine California counties. International Journal of Public Health, 55, 113-121. http://dx.doi.org/10.1007/ s00038-009-0076-0

58. CDC, 2015: Emergency Preparedness and Response: Extreme Heat. Centers for Disease Control and Prevention, Atlanta, GA. http://www.bt.cdc.gov/disasters/extremeheat/

68. Ostro, B., S. Rauch, R. Green, B. Malig, and R. Basu, 2010: The effects of temperature and use of air conditioning on hospitalizations. American Journal of Epidemiology, 172, 1053-1061. http://dx.doi.org/10.1093/aje/kwq231

59. EPA, 2015: Natural Disasters: Extreme Heat. U.S. Environmental Protection Agency, Washington, D.C. http://epa. gov/naturaldisasters/extremeheat.html 60. Fischer, E.M. and R. Knutti, 2015: Anthropogenic contribution to global occurrence of heavy-precipitation and high-temperature extremes. Nature Climate Change, 5, 560564. http://dx.doi.org/10.1038/nclimate2617 61. Davis, R.E., P.C. Knappenberger, W.M. Novicoff, and P.J. Michaels, 2003: Decadal changes in summer mortality in US cities. International Journal of Biometeorology, 47, 166175. http://dx.doi.org/10.1007/s00484-003-0160-8


69. Schwartz, J., J.M. Samet, and J.A. Patz, 2004: Hospital admissions for heart disease: The effects of temperature and humidity. Epidemiology, 15, 755-761. http://dx.doi. org/10.1097/01.ede.0000134875.15919.0f 70. Turner, L.R., A.G. Barnett, D. Connell, and S. Tong, 2012: Ambient temperature and cardiorespiratory morbidity: A systematic review and meta-analysis. Epidemiology, 23, 594606. http://dx.doi.org/10.1097/EDE.0b013e3182572795 71. Fletcher, B.A., S. Lin, E.F. Fitzgerald, and S.A. Hwang, 2012: Association of summer temperatures with hospital admissions for renal diseases in New York State: A case-crossover study. American Journal of Epidemiology, 175, 907-916. http://dx.doi.org/10.1093/aje/kwr417

64



72. Li, B., S. Sain, L.O. Mearns, H.A. Anderson, S. Kovats, K.L. Ebi, M.Y.V. Bekkedal, M.S. Kanarek, and J.A. Patz, 2012: The impact of extreme heat on morbidity in Milwaukee, Wisconsin. Climatic Change, 110, 959-976. http://dx.doi. org/10.1007/s10584-011-0120-y

80. Peng, R.D., J.F. Bobb, C. Tebaldi, L. McDaniel, M.L. Bell, and F. Dominici, 2011: Toward a quantitative estimate of future heat wave mortality under global climate change. Environmental Health Perspectives, 119, 701-706. http://dx. doi.org/10.1289/ehp.1002430

73. Alessandrini, E., S. Zauli Sajani, F. Scotto, R. Miglio, S. Marchesi, and P. Lauriola, 2011: Emergency ambulance dispatches and apparent temperature: A time series analysis in Emilia–Romagna, Italy. Environmental Research, 111, 11921200. http://dx.doi.org/10.1016/j.envres.2011.07.005

81. Knowlton, K., B. Lynn, R.A. Goldberg, C. Rosenzweig, C. Hogrefe, J.K. Rosenthal, and P.L. Kinney, 2007: Projecting heat-related mortality impacts under a changing climate in the New York City region. American Journal of Public Health, 97, 2028-2034. http://dx.doi.org/10.2105/ Ajph.2006.102947

74. Huang, C., A.G. Barnett, X. Wang, P. Vaneckova, G. FitzGerald, and S. Tong, 2011: Projecting future heat-related mortality under climate change scenarios: A systematic review. Environmental Health Perspectives, 119, 1681-1690. http://dx.doi.org/10.1289/Ehp.1103456 75. Barreca, A.I., 2012: Climate change, humidity, and mortality in the United States. Journal of Environmental Economics and Management, 63, 19-34. http://dx.doi.org/10.1016/j. jeem.2011.07.004 76. Greene, S., L.S. Kalkstein, D.M. Mills, and J. Samenow, 2011: An examination of climate change on extreme heat events and climate–mortality relationships in large U.S. cities. Weather, Climate, and Society, 3, 281-292. http://dx.doi. org/10.1175/WCAS-D-11-00055.1 77. Honda, Y., M. Kondo, G. McGregor, H. Kim, Y.-L. Guo, Y. Hijioka, M. Yoshikawa, K. Oka, S. Takano, S. Hales, and R.S. Kovats, 2014: Heat-related mortality risk model for climate change impact projection. Environmental Health and Preventive Medicine, 19, 56-63. http://dx.doi.org/10.1007/ s12199-013-0354-6 78. Voorhees, A.S., N. Fann, C. Fulcher, P. Dolwick, B. Hubbell, B. Bierwagen, and P. Morefield, 2011: Climate change-related temperature impacts on warm season heat mortality: A proof-of-concept methodology using BenMAP. Environmental Science & Technology, 45, 1450-1457. http://dx.doi. org/10.1021/es102820y 79. Wu, J., Y. Zhou, Y. Gao, J.S. Fu, B.A. Johnson, C. Huang, Y.-M. Kim, and Y. Liu, 2014: Estimation and uncertainty analysis of impacts of future heat waves on mortality in the eastern United States. Environmental Health Perspectives, 122, 10-16. http://dx.doi.org/10.1289/ehp.1306670


82. Petkova, E.P., R.M. Horton, D.A. Bader, and P.L. Kinney, 2013: Projected heat-related mortality in the U.S. urban northeast. International Journal of Environmental Research and Public Health, 10, 6734-6747. http://dx.doi. org/10.3390/ijerph10126734 83. Isaksen, T.B., M. Yost, E. Hom, and R. Fenske, 2014: Projected health impacts of heat events in Washington State associated with climate change. Reviews on Environmental Health, 29, 119-123. http://dx.doi.org/10.1515/reveh-2014-0029 84. Jackson, J.E., M.G. Yost, C. Karr, C. Fitzpatrick, B.K. Lamb, S.H. Chung, J. Chen, J. Avise, R.A. Rosenblatt, and R.A. Fenske, 2010: Public health impacts of climate change in Washington State: Projected mortality risks due to heat events and air pollution. Climatic Change, 102, 159-186. http://dx.doi.org/10.1007/s10584-010-9852-3 85. Ostro, B., S. Rauch, and S. Green, 2011: Quantifying the health impacts of future changes in temperature in California. Environmental Research, 111, 1258-1264. http://dx.doi. org/10.1016/j.envres.2011.08.013 86. Petkova, E.P., D.A. Bader, G.B. Anderson, R.M. Horton, K. Knowlton, and P.L. Kinney, 2014: Heat-related mortality in a warming climate: Projections for 12 U.S. cities. International Journal of Environmental Research and Public Health, 11, 11371-11383. http://dx.doi.org/10.3390/ijerph111111371 87. Lin, S., W.-H. Hsu, A.R. Van Zutphen, S. Saha, G. Luber, and S.-A. Hwang, 2012: Excessive heat and respiratory hospitalizations in New York State: Estimating current and future public health burden related to climate change. Environmental Health Perspectives, 120, 1571-1577. http://dx. doi.org/10.1289/ehp.1104728

65



88. Brikowski, T.H., Y. Lotan, and M.S. Pearle, 2008: Climate-related increase in the prevalence of urolithiasis in the United States. Proceedings of the National Academy of Sciences, 105, 9841-9846. http://dx.doi.org/10.1073/pnas.0709652105

98. Kovats, R.S. and S. Hajat, 2008: Heat stress and public health: A critical review. Annual Review of Public Health, 29, 41-55. http://dx.doi.org/10.1146/annurev. publhealth.29.020907.090843

89. Fakheri, R.J. and D.S. Goldfarb, 2011: Ambient temperature as a contributor to kidney stone formation: Implications of global warming. Kidney International, 79, 1178-1185. http://dx.doi.org/10.1038/ki.2011.76

99. Gamble, J.L., B.J. Hurley, P.A. Schultz, W.S. Jaglom, N. Krishnan, and M. Harris, 2013: Climate change and older Americans: State of the science. Environmental Health Perspectives, 121, 15-22. http://dx.doi.org/10.1289/ ehp.1205223

90. Tasian, G.E., J.E. Pulido, A. Gasparrini, C.S. Saigal, B.P. Horton, J.R. Landis, R. Madison, and R. Keren, 2014: Daily mean temperature and clinical kidney stone presentation in five U.S. metropolitan areas: A time-series analysis. Environmental Health Perspectives, 122, 1081-1087. http://dx.doi. org/10.1289/ehp.1307703 91. Watkiss, P. and A. Hunt, 2012: Projection of economic impacts of climate change in sectors of Europe based on bottom up analysis: Human health. Climatic Change, 112, 101126. http://dx.doi.org/10.1007/s10584-011-0342-z 92. Mercer, J.B., 2003: Cold—an underrated risk factor for health. Environmental Research, 92, 8-13. http://dx.doi. org/10.1016/s0013-9351(02)00009-9 93. Ebi, K.L. and D. Mills, 2013: Winter mortality in a warming climate: A reassessment. Wiley Interdisciplinary Reviews: Climate Change, 4, 203-212. http://dx.doi.org/10.1002/ wcc.211 94. Kinney, P.L., J. Schwartz, M. Pascal, E. Petkova, A. Le Tertre, S. Medina, and R. Vautard, 2015: Winter season mortality: Will climate warming bring benefits? Environmental Research Letters, 10, 064016. http://dx.doi.org/10.1088/17489326/10/6/064016 95. Davis, R.E., P.C. Knappenberger, W.M. Novicoff, and P.J. Michaels, 2002: Decadal changes in heat-related human mortality in the eastern United States. Climate Research, 22, 175-184. http://dx.doi.org/10.3354/cr022175 96. Ebi, K.L., T.J. Teisberg, L.S. Kalkstein, L. Robinson, and R.F. Weiher, 2004: Heat watch/warning systems save lives: Estimated costs and benefits for Philadelphia 1995–98. Bulletin of the American Meteorological Society, 85, 1067-1073. http://dx.doi.org/10.1175/bams-85-8-1067 97. Sheridan, S.C., A.J. Kalkstein, and L.S. Kalkstein, 2009: Trends in heat-related mortality in the United States, 1975–2004. Natural Hazards, 50, 145-160. http://dx.doi. org/10.1007/s11069-008-9327-2


100. Balbus, J.M. and C. Malina, 2009: Identifying vulnerable subpopulations for climate change health effects in the United States. Journal of Occupational and Environmental Medicine, 51, 33-37. http://dx.doi.org/10.1097/ JOM.0b013e318193e12e 101. Benmarhnia, T., S. Deguen, J.S. Kaufman, and A. Smargiassi, 2015: Review article: Vulnerability to heat-related mortality: A systematic review, meta-analysis, and meta-regression analysis. Epidemiology, 26, 781-793. http://dx.doi. org/10.1097/EDE.0000000000000375 102. CDC, 2011: Extreme Heat and Your Health: Heat and Infants and Children. Centers for Disease Control and Prevention, Atlanta, GA. http://www.cdc.gov/extremeheat/children.html 103. Wasilevich, E.A., F. Rabito, J. Lefante, and E. Johnson, 2012: Short-term outdoor temperature change and emergency department visits for asthma among children: A case-crossover study. American Journal of Epidemiology, 176, S123-S130. http://dx.doi.org/10.1093/aje/kws326 104. Kerr, Z.Y., D.J. Casa, S.W. Marshall, and R.D. Comstock, 2013: Epidemiology of exertional heat illness among U.S. high school athletes. American Journal of Preventive Medicine, 44, 8-14. http://dx.doi.org/10.1016/j.amepre.2012.09.058 105. Basu, R., B. Malig, and B. Ostro, 2010: High ambient temperature and the risk of preterm delivery. American Journal of Epidemiology, 172, 1108-1117. http://dx.doi.org/10.1093/ aje/kwq170 106. Carolan-Olah, M. and D. Frankowska, 2014: High environmental temperature and preterm birth: A review of the evidence. Midwifery, 30, 50-59. http://dx.doi.org/10.1016/j. midw.2013.01.011 107. Beltran, A.J., J. Wu, and O. Laurent, 2014: Associations of meteorology with adverse pregnancy outcomes: A systematic review of preeclampsia, preterm birth and birth weight. International Journal of Environmental Research and Public Health, 11, 91-172. http://dx.doi.org/10.3390/ijerph110100091

66



108. Uejio, C.K., O.V. Wilhelmi, J.S. Golden, D.M. Mills, S.P. Gulino, and J.P. Samenow, 2011: Intra-urban societal vulnerability to extreme heat: The role of heat exposure and the built environment, socioeconomics, and neighborhood stability. Health & Place, 17, 498-507. http://dx.doi. org/10.1016/j.healthplace.2010.12.005 109. Ramin, B. and T. Svoboda, 2009: Health of the homeless and climate change. Journal of Urban Health, 86, 654-664. http://dx.doi.org/10.1007/s11524-009-9354-7 110. Gronlund, C.J., 2014: Racial and socioeconomic disparities in heat-related health effects and their mechanisms: A review. Current Epidemiology Reports, 1, 165-173. http://dx.doi. org/10.1007/s40471-014-0014-4 111. Hansen, A., L. Bi, A. Saniotis, and M. Nitschke, 2013: Vulnerability to extreme heat and climate change: Is ethnicity a factor? Global Health Action, 6. http://dx.doi.org/10.3402/ gha.v6i0.21364 112. O’Neill, M.S., A. Zanobetti, and J. Schwartz, 2005: Disparities by race in heat-related mortality in four US cities: The role of air conditioning prevalence. Journal of Urban Health, 82, 191-197. http://dx.doi.org/10.1093/jurban/jti043 113. Arbury, S., B. Jacklitsch, O. Farquah, M. Hodgson, G. Lamson, H. Martin, and A. Profitt, 2014: Heat illness and death among workers – United States, 2012-2013. MMWR. Morbidity and Mortality Weekly Report, 63, 661-665. http:// www.cdc.gov/mmwr/preview/mmwrhtml/mm6331a1.htm 114. Lundgren, K., K. Kuklane, C. Gao, and I. Holmer, 2013: Effects of heat stress on working populations when facing climate change. Industrial Health, 51, 3-15. http:// dx.doi.org/10.2486/indhealth.2012-0089 :// WOS:000314383700002 115. Hansen, A., P. Bi, M. Nitschke, P. Ryan, D. Pisaniello, and G. Tucker, 2008: The effect of heat waves on mental health in a temperate Australian city. Environmental Health Perspectives, 116, 1369-1375. http://dx.doi.org/10.1289/ehp.11339 116. Martin-Latry, K., M.P. Goumy, P. Latry, C. Gabinski, B. Bégaud, I. Faure, and H. Verdoux, 2007: Psychotropic drugs use and risk of heat-related hospitalisation. European Psychiatry, 22, 335-338. http://dx.doi.org/10.1016/j.eurpsy.2007.03.007


117. Page, L.A., S. Hajat, R.S. Kovats, and L.M. Howard, 2012: Temperature-related deaths in people with psychosis, dementia and substance misuse. The British Journal of Psychiatry, 200, 485-490. http://dx.doi.org/10.1192/bjp. bp.111.100404 118. Ranson, M., 2014: Crime, weather, and climate change. Journal of Environmental Economics and Management, 67, 274-302. http://dx.doi.org/10.1016/j.jeem.2013.11.008 119. Dunne, J.P., R.J. Stouffer, and J.G. John, 2013: Reductions in labour capacity from heat stress under climate warming. Nature Climate Change, 3, 563-566. http://dx.doi. org/10.1038/nclimate1827 120. Graff Zivin, J. and M. Neidell, 2014: Temperature and the allocation of time: Implications for climate change. Journal of Labor Economics, 32, 1-26. http://dx.doi. org/10.1086/671766 121. Sheridan, S.C. and T.J. Dolney, 2003: Heat, mortality, and level of urbanization: Measuring vulnerability across Ohio, USA. Climate Research, 24, 255-265. http://dx.doi. org/10.3354/cr024255 122. Orlowsky, B. and S.I. Seneviratne, 2012: Global changes in extreme events: Regional and seasonal dimension. Climatic Change, 10, 669-696. http://dx.doi.org/10.1007/s10584011-0122-9 123. Robeson, S.M., C.J. Willmott, and P.D. Jones, 2014: Trends in hemispheric warm and cold anomalies. Geophysical Research Letters, 41, 9065-9071. http://dx.doi. org/10.1002/2014gl062323 124. Liu, J., M. Song, R.M. Horton, and Y. Hu, 2013: Reducing spread in climate model projections of a September ice-free Arctic. Proceedings of the National Academy of Sciences, 110, 12571-12576. http://dx.doi.org/10.1073/ pnas.1219716110 125. Liu, J., J.A. Curry, H. Wang, M. Song, and R.M. Horton, 2012: Impact of declining Arctic sea ice on winter snowfall. Proceedings of the National Academy of Sciences, 109, 40744079. http://dx.doi.org/10.1073/pnas.1114910109 126. Francis, J.A. and S.J. Vavrus, 2012: Evidence linking Arctic amplification to extreme weather in mid-latitudes. Geophysical Research Letters, 39, L06801. http://dx.doi. org/10.1029/2012GL051000

67



End

127. Barnes, E.A., 2013: Revisiting the evidence linking Arctic amplification to extreme weather in midlatitudes. Geophysical Research Letters, 40, 4734-4739. http://dx.doi.org/10.1002/ grl.50880 128. Wallace, J.M., I.M. Held, D.W.J. Thompson, K.E. Trenberth, and J.E. Walsh, 2014: Global warming and winter weather. Science, 343, 729-730. http://dx.doi.org/10.1126/ science.343.6172.729 129. Gosling, S.N., G.R. McGregor, and J.A. Lowe, 2009: Climate change and heat-related mortality in six cities Part 2: Climate model evaluation and projected impacts from changes in the mean and variability of temperature with climate change. International Journal of Biometeorology, 53, 31-51. http://dx.doi.org/10.1007/s00484-008-0189-9


68
